Spring 2008 6.831 User Interface Design and Implementation 1Lecture 18: Predictive Evaluation UI Hall of Fame or Shame?Spring 2008 6.831 User Interface Design and Implementation 2 From Daniel Gutierrez: “The speed dial implemented into the Opera web browser (screen shot attached) is designed to provide a type of visual bookmark for quick access to web pages. It appears in a newly created tab and begins to load each of the websites to create a preview thumbnail for each one. When you click on any of these, it navigates the browser to the web page. One cool thing about this is that it actually starts to cache images and similar content from the websites it’s loading while the browser is idling on this screen.” Let’s talk about this interface in terms of: - visibility - efficiency - graphic designToday’s Topics• Keystroke-level models•GOMS• CPM-GOMSSpring 2008 6.831 User Interface Design and Implementation 3 Today’s lecture is about predictive evaluation – the holy grail of usability engineering. If we had an accurate model for the way a human used a computer interface, we would be able to predict the usability of a design, without having to actually build it, test it against real people, and measure their behavior. User interface design would then become more like other fields of engineering. Civil engineers can use models (of material stress and strain) to predict the load that can be carried by a bridge; they don’t have to build it and test it to destruction first. As user interface designers, we’d like to do the same thing. Predictive Evaluation• Predictive evaluation uses an engineering model of human cognition to predict usability• Model is–abstract– quantitative–approximate– estimated from user experimentsSpring 2008 6.831 User Interface Design and Implementation 4PP CPMP~100 ms ~70 ms ~70 ms At its heart, any predictive evaluation technique requires a model for how a user interacts with an interface. We’ve already seen one such model, the Newell/Card/Moran human information processing model. This model needs to be abstract – it can’t be as detailed as an actual human being (with billions of neurons, muscles, and sensory cells), because it wouldn’t be practical to use for prediction. The model we looked at boiled down the rich aspects of information processing into just three processors and two memories. It also has to be quantitative, i.e., assigning numerical parameters to each component. Without parameters, we won’t be able to compute a prediction. We might still be able to do qualitative comparisons, such as we’ve already done to compare, say, Mac menu bars with Windows menu bars, or cascading submenus. But our goals for predictive evaluation are more ambitious. These numerical parameters are necessarily approximate; first because the abstraction in the model aggregates over a rich variety of different conditions and tasks; and second because human beings exhibit large individual differences, sometimes up to a factor of 10 between the worst and the best. So the parameters we use will be averages, and we may want to take the variance of the parameters into account when we do calculations with the model.Where do the parameters come from? They’re estimated from experiments with real users. The numbers seen here for the general model of human information processing (e.g., cycle times of processors and capacities of memories) were inferred from a long literature of cognitive psychology experiments. But for more specific models, parameters may actually be estimated by setting up new experiments designed to measure just that parameter of the model. Advantages of Predictive Evaluation•Don’t have to build UI prototype– Can compare design alternatives with no implementation whatsoever•Don’t have to test real live users• Theory provides explanations of UI problems– So it points to the areas where design can be improved– User testing may only reveal problems, not explain themSpring 2008 6.831 User Interface Design and Implementation 5 Predictive evaluation doesn’t need real users (once the parameters of the model have been estimated, that is). Not only that, but predictive evaluation doesn’t even need a prototype. Designs can be compared and evaluated without even producing design sketches or paper prototypes, let alone code. Another key advantage is that the predictive evaluation not only identifies usability problems, but actually provides an explanation of them based on the theoretical model underlying the evaluation. So it’s much better at pointing to solutions to the problems than either inspection techniques or user testing. User testing might show that design A is 25% slower than design B at a doing a particular task, but it won’t explain why. Predictive evaluation breaks down the user’s behavior into little pieces, so that you can actually point at the part of the task that was slower, and see why it was slower.Keystroke-Level Model (KLM)• Keystroke• Button press or release with mouse• Point with mouse• Draw line with mouse• Home hands between mouse and keyboard• Mentally prepareSpring 2008 6.831 User Interface Design and Implementation 6 The first predictive model was the keystroke level model (proposed by Card, Moran & Newell, “The Keystroke Level Model for User Performance Time with Interactive Systems”, CACM, v23 n7, July 1978). This model seeks to predict efficiency (time taken by expert users doing routine tasks) by breaking down the user’s behavior into a sequence of the five primitive operators shown here. Most of the operators are physical – the user is actually moving their muscles to perform them. The M operator is different – it’s purely mental (which is somewhat problematic, because it’s hard to observe and estimate). The M operator stands in for any mental operations that the user does. M operators separate the task into chunks, or steps, and represent the time needed for the user to recall the next step from long-term memory. KLM Analysis• Encode a method as a sequence of physical operators (KPHD)• Use heuristic rules to insert mental operators (M)• Add up times for each operator to get total time for methodSpring 2008 6.831 User Interface Design and Implementation 7 Here’s how to create a keystroke level model for a task. First, you have to focus on a particular method for doing the task. Suppose the task is deleting a
View Full Document
Unlocking...